Deepikachintamreddy commited on
Commit ·
c8197e5
1
Parent(s): 991b3a8
fix: model defaults 0.01 not 0.0
Browse files- server/models.py +5 -7
server/models.py
CHANGED
|
@@ -12,7 +12,7 @@ class ConfigDebugObservation(Observation):
|
|
| 12 |
Inherits done: bool, reward: Optional[float], metadata from Observation.
|
| 13 |
"""
|
| 14 |
broken_config: str = ""
|
| 15 |
-
ground_truth: str = ""
|
| 16 |
file_type: str = ""
|
| 17 |
error_message: str = ""
|
| 18 |
task_id: str = ""
|
|
@@ -20,7 +20,7 @@ class ConfigDebugObservation(Observation):
|
|
| 20 |
difficulty: str = ""
|
| 21 |
num_bugs: int = 0
|
| 22 |
bugs_found_so_far: int = 0
|
| 23 |
-
previous_reward: float = 0.
|
| 24 |
|
| 25 |
|
| 26 |
class ConfigDebugState(State):
|
|
@@ -30,13 +30,11 @@ class ConfigDebugState(State):
|
|
| 30 |
current_task_id: str = ""
|
| 31 |
current_step: int = 0
|
| 32 |
max_steps: int = 5
|
| 33 |
-
total_reward: float = 0.
|
| 34 |
is_done: bool = False
|
| 35 |
tasks_completed: List[str] = []
|
| 36 |
tasks_remaining: List[str] = []
|
| 37 |
-
|
| 38 |
-
# Enhanced state fields for better RL signal
|
| 39 |
bugs_found_so_far: int = 0
|
| 40 |
current_error_message: Optional[str] = None
|
| 41 |
-
progress_ratio: float = 0.
|
| 42 |
-
current_difficulty: Optional[str] = None
|
|
|
|
| 12 |
Inherits done: bool, reward: Optional[float], metadata from Observation.
|
| 13 |
"""
|
| 14 |
broken_config: str = ""
|
| 15 |
+
ground_truth: str = ""
|
| 16 |
file_type: str = ""
|
| 17 |
error_message: str = ""
|
| 18 |
task_id: str = ""
|
|
|
|
| 20 |
difficulty: str = ""
|
| 21 |
num_bugs: int = 0
|
| 22 |
bugs_found_so_far: int = 0
|
| 23 |
+
previous_reward: float = 0.01
|
| 24 |
|
| 25 |
|
| 26 |
class ConfigDebugState(State):
|
|
|
|
| 30 |
current_task_id: str = ""
|
| 31 |
current_step: int = 0
|
| 32 |
max_steps: int = 5
|
| 33 |
+
total_reward: float = 0.01
|
| 34 |
is_done: bool = False
|
| 35 |
tasks_completed: List[str] = []
|
| 36 |
tasks_remaining: List[str] = []
|
|
|
|
|
|
|
| 37 |
bugs_found_so_far: int = 0
|
| 38 |
current_error_message: Optional[str] = None
|
| 39 |
+
progress_ratio: float = 0.01
|
| 40 |
+
current_difficulty: Optional[str] = None
|